Cut, 1989, 30, 406-415

Progress report

Non-cirrhotic portal fibrosis

Background and nomenclature

Splenomegaly with or without anaemia has drawn the attention of a numberof workers since Banti of North Italy described in 1889, a diseasecharacterised by splenomegaly and secondary liver cirrhosis. 'Over the yearsit became clear that Banti's syndrome, as it used to be called, is an ill defineddisorder, consisting of a variety of diseases that present with splenomegaly,hypersplenism and portal hypertension.

It was in 1956, that some physicians in India first drew attention to aclinical syndrome of portal hypertension distinct from cirrhosis. In 1962,Ramalingaswami and colleagues after a careful study of the autopsy materialof some of these patients, reported an entity of splenomegaly with non-cirrhotic liver disease in North Indian patients.' Soon, many Indian workersreaffirmed the separate identity of this condition.74 In 1969, a workshop wasorganised by the Indian Council of Medical Research to review the availableinformation on this condition. The study group decided to name thisclinicopathological entity 'non-cirrhotic portal fibrosis' (NCPF).' Boyerwhile working in Calcutta, saw almost identical patients and reported themunder the term 'idiopathic portal hypertension'.6 Several workers at nearlythe same time, described patients with portal hypertension with a patentportal vein and without evidence of cirrhosis.7 Mikkelsen et al identified 36patients with portal hypertension without cirrhosis, who had phlebosclerosisof intra and extrahepatic portal veins and called this condition as 'hepato-portal sclerosis'.' During the following years, The International Associationfor the Study of Liver, recognised a group of patients with 'portal and septalfibrosis associated with portal hypertension occurring predominantly inadults, mostly in males, in tropical zones' and accepted the term 'hepato-portal sclerosis'." It is, however, important to remember that Mikkelsen'spatient material described under the same term, included a significantnumber of patients with partial or complete portal vein block,s a conditionwhich is better known as extra hepatic portal vein obstruction (EHPVO).Whatever may be the terminology applied, it is now abundantly clear that

a distinct syndrome of portal hypertension without any evidence of cirrhosisof liver and EHPVO exists and is seen in many parts of the world. There are,however, subtle differences between the patients reported from differentareas. In an excellent review, Okuda recently provided a detailed account ofpatients with idiopathic portal hypertension (IPH) seen in Japan."' In thisreview, NCPF is described as it is understood in India and is compared withIPH and similarly described entities.

Epidemiology

Non-cirrhotic portal fibrosis has been reported from almost all parts of

406

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

Non-cirrhotic porttalfibrosis

India." '4 Nearly 25-30% of all patients with portal hypertension in Indiawho undergo surgery or sclerotherapy, have NCPF.'-3 '4 In fact, the largestnumber of cases of NCPF (or a similar disease elsewhere) in the world havebeen reported from India. No large epidemiological studies are available,however, to determine the incidence of NCPF in general population. Themajority of NCPF patients come from a lower or middle socio-economicbackground.

Non-cirrhotic portal fibrosis is usually a disease of young Indian men.Except for a single report from Chandigarh of female predisposition,'" allseries describe a male predominance (M:F, 2:1 to 4: 1 ) . '1" '3 "' This is incontrast with patients of IPH from Japan" and USA'" where M:F ratio isabout 1:3.The mean age of NCPF patients has been reported to vary from 25

to 35 years, about two decades younger than IPH patients.'"." Some workershave described a double peak of onset of NCPF, one at 21-25 and anotherat 36-40 years." This finding has, however, not been confirmed byothers.

Aetiology

Despite its common occurrence, the aetiopathogenesis of NCPF is poorlyunderstood. A number of hypotheses have been put forward in the past,based mainly on anecdotal experiences and case reports.

EXPOSURE TO TRACF MFTAILS AND CHEMICAL SClinical association between ingestion of certain trace metals and hepaticfibrosis was observed long ago. Chronic ingestion of arsenic has beenincriminated in the causation of NCPF because it causes predominantfibrosis of intrahepatic portal veins.'7 Datta et al had reported significantlyhigher hepatic arsenic concentrations in NCPF patients compared withcontrols.'" Similar levels were, however, subsequently seen in patients withcirrhosis and Indian childhood cirrhosis."' In seven of our NCPF patientsstudied by electron probe microanalysis, hepatic arsenic content was notdifferent from controls.The possibility of ingestion of indigenous drugs, food adulterants, herbal

and fungal toxins or parasitic infection has also been raised with littlesubstantiation."'

INFECTIVE AGENTSNon-cirrhotic portal fibrosis is predominantly a disease of the poor and iscommon in countries like India, where the incidence of umbilical sepsis,diarrhoeal disease and bacterial infections is very high. It is speculated thatthrombophlebitis and repeated embolisation into the portal circulationbecause of any abdominal infection initiated at birth or at a later age, couldresult in NCPF.' Idiopathic portal hypertension like changes have beenreported after injecting dead non-pathogenic colon bacilli into the portalvein of rabbits, previously sensitised with the same bacteria.2"A higher incidence of malarial antibody positivity has been reported in

NCPF patients compared with cirrhotics and controls.2' Although malariais endemic in India and splenomegaly is common to both diseases, no directassociation has so far been found between the two conditions.

407

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

IMMUNOI OGIC AND IMMUNOGENE[IC HYPOTHESESEvidences supporting these hypotheses include - (i) occurrence of non-compressive myelopathy after lieno-renal shunt in a patient with NCPF,'-(ii) a reduction in suppressor/cytotoxic lymphocytes (T8) and an alterationin the helper/suppressor (T4/T8) lymphocytes in NCPF,'4 (iii) a reduction inthe cell mediated immune status in NCPF patients," and (iv) someresemblence with schistosomiasis- a disease known to result from theimmunological injury caused by the eggs lodged in the radicles of portalvein.2" Unfortunately, whether these abnormalities are the cause or theeffect of the disease is not known.

Familial aggregation has been recently reported in NCPF.7 In anotherstudy of 48 NCPF patients, the frequency of HLA-DR3 was found to besignificantly higher than controls. The results suggested an autoimmunebasis of the disease and that susceptibility to NCPF may be HLA class I1-mediated.>

INCREASED Bl OOD FLOW

Since the initial description of Banti, disproportionate splenomegaly inthese patients has intrigued the investigators. Majority of the workers havefound increased splenic and hepatic blood flow in NCPF patients" 2' butwhether this is the primary event leading to portal hypertension or resultsfrom congestion caused by increased portal vascular resistance, is not clear.

Pathology

GROSSThe liver may be normal to markedly nodular. "'1 Although, the surface mayresemble cirrhosis, the deeper portions of liver do not show any nodularityand the liver may weigh from less to more than normal."33' On the cutsurface, hypertrophic nodules may occasionally be seen, compressing theportal vein.3' The portal venous system shows prominent and dilatedbranches with marked sclerosis of the walls. In autopsy reports, frequentthrombosis of the large and small portal vein branches with accompanyingareas of ischaemic necrosis has been reported.' Significant perivascularfibrosis along the portal vein and its branches is an important feature.

MICROSCOPY

Histological lesions in the liver were aptly summarised by Nayak et alas, 'obliterative portal venopathy of liver', which indicates markedsubendothelial thickening of the large and medium sized intrahepaticbranches of the portal vein with a patchy segmental distribution..' In thesections obtained from autopsy livers, organised thrombi with recanalisa-tion, scarring of the terminal portal tracts with obliteration or disappearanceof the portal vein radicles and appearance of aberrant vessels are oftenseen. The lobular architecture and hepatic parenchyma is unaffected. In theneedle biopsy specimens, the changes are often minimal or absent.Although, Okuda et al maintain that the pathology of NCPF and IPH isidentical,33 subtle difference are noteworthy (Table).

U l TRASTRUCTU RIA widening of the intercellular and Disse's spaces with fibrogenesis in the

408 Szaritz

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

Non-cirrhotic portalfibrosis

Table Comparison ofsome features ofNCPF and IPH

Feature NCPF IPH

Male:female 2 to 4:1 1:3Mean age (yr) 25 to 35 43 to 56Clinical features"' 1'39Haematemesis/malena 94% 4(0%Mass in left upper abdomen 6.0% 40%Ascites 201% 10%Anaemia and symptoms related to it - 35%Jaundice 2 1%oBleeding tendency - 6%Splenomegaly 95% 880%Hepatomegaly 61% 41 0/o

Pathology3'-l'Irregular parenchymal nodules 53% 29%Nodules on postero-inferior surface 25% 40°.Attempted pseudolobule 72% 300/0Bile duct proliferation 38% 40/o

Haemodynamics'3 ,5Wedged hepatic vein pressure Normal or raised Always raised

perisinusoidal space and occasional capillarisation of the sinusoids isseen.'3 34 The cell membrane between the hepatocytes shows development ofmicrovilli. The intracytoplasmic components of hepatocytes are normal.

It is not clear whether the morphological changes seen in NCPF aresecondary to circulatory insufficiency, as argued by Japanese workers, asthey are not seen in patients with extrahepatic portal vein obstruction.

HAEMODYNAMICSThe intrasplenic and portal vein pressures are markedly elevated in patientswith NCPF.6"' '3 One of the main controversies in the hemodynamicevaluation of NCPF concerns wedged hepatic vein pressure (WHVP)measurements. Wedged hepatic vein pressure has been reported fromnormal to moderately raised.6' '" Sama et al found increased WHVP in91% patients.'3 We observed it to be 8 mm or less in nearly half the patientsand raised in the rest.35 The differences in various studies could partly bebecause of the differences in the extent and site of fibrosis in the liver ofpatients. In patients with IPH on the other hand, WHVP is almost alwaysraised (Table).'5The fact that WHVP is lower than portal pressure indicates a resistance to

portal blood flow (portohepatic gradient). Until recently, the exact site ofthe resistance was not clear. It has now been shown that two independentpressure gradients exist in NCPF patients, the first between intrasplenic andintrahepatic pressure and the second, between intrahepatic and WHVP.35 37While demonstration of the former gradient haemodynamically outlines thepresence of presinusoidal block, the existence of the later gradient, reflectsa perisinusoidal resistance to flow of portal blood. As both wedged andintrahepatic pressures underestimate the portal pressure in NCPF,intravariceal pressure should be used as a reliable indicator of portalpressure.Two dimensional contrast echocardiography has been used to study the

pulmonary arteriovenous shunts and cardiac chamber abnormalities inNCPF. While shunts were not detected in any patient, left atrial and

409

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

ventricular enlargement was seen in 64 and 57% and reduced left ventricularejection fraction in 36% patients with NCPF.3" Similar changes were alsoseen in cirrhotics suggesting that hyperdynamic portal hypertensive state isprobably responsible for these changes. Quantitative evaluation of intra andextrahepatic portosystemic shunting has not been carried out for NCPFpatients as has been done for IPH patients.""5Although the haemodynamic investigations have helped to improve our

understanding of the pathophysiology of NCPF, they have failed to providean insight into the cause of portal hypertension and the initial events leadingto development of increased blood flow, splenomegaly and pre andperisinusoidal resistance to portal blood flow.

Clinical profile

Patients with NCPF are usually young and look well. They almost alwayspresent with one or more episodes of haematemesis, which are welltolerated, and with a long standing history of a mass in the left hypo-chondrium. Jaundice and signs of liver failure are rare and occur transientlyafter haematemesis or during terminal illness.'339 Except for an initialreport, it is believed that ascites is uncommon in NCPF and if present, is mildand develops after a bleed."" These features of NCPF are somewhatdifferent from those reported in the National Study on IPH in Japan,"' wherea mass in the left hypochondrium, anaemia and symptoms related toanaemia were almost as common a presentation as haematemesis. Mostconspicuous physical finding in NCPF is splenomegaly which at times can fillmost of the abdomen. Overt or subclinical encephalopathy is not a feature ofNCPF" except after shunt surgery.'34'

Laboratory investigations

Anaemia is seen in majority of patients. It is usually of normocyticnormochromic or normocytic hypochromic type. Leucopenia (<5000/cmm)and thrombocytopenia (<100 000/cmm) are also frequently seen.6'3 Thesechanges in part could be a result of increase in plasma volume and splanchnicpooling of blood.6 Prothrombin time and coagulation parameters are usuallywithin normal limits. Bone marrow is hypercellular. Although RBC survivalhas not been studied in detail in NCPF, the incidence of hypersplenism is notas high as that reported in IPH.'9

Tests of liver function like serum bilirubin, globulin, alkaline phospha-tase and transaminases (ALT and AST) are normal or mildly deranged.Bromsulphalein retention was reported to be abnormal (> 16% retention) in20% patients.'- An increase in arterial ammonia concentration4" and areduction in pentagastrin stimulated maximum gastric acid secretion43 hasbeen shown. The biliary lipid composition is reported to be normal in NCPFpatients.'

Radiological features

Oesophageal varices were shown by barium swallow in 68% patients."5 In 78of the 81 (96%) NCPF patients studied by us with UGI endoscopy,oesophageal varices were present.

Sarin410

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

Non-cirrhoticportalfibrosis

Splenoportovenography (SPV) has been the common and often the onlymode of investigation used to evaluate the splenoportal axis in NCPFpatients in India."II '4 Splenoportovenography usually shows dilatation ofthe portal and splenic veins along with various collaterals. Some workershave stressed the importance of selective dilatation of the left branch ofportal vein with sudden narrowing ('cut off sign) of major intrahepaticbranches in NCPF.' 1- These features are, however, not diagnostic.There are no published data on the sonographic findings in NCPF. In our

initial experience in 36 patients, marked thickening of the portal vein wallsspecially of intrahepatic branches, was observed. In three patients, athrombus was seen in the right or left main branch of the portal vein.

Relationship between NCPF and extrahepatic portal vein obstruction(EHPVO)

Partial or complete obstruction of the extrahepatic portal vein is one of thethree important causes of portal hypertension in India; the other two beingcirrhosis and NCPF. Extrahepatic portal vein obstruction is usually seen inyoung male children who often belong to poor social class. Although thehistory of umbilical sepsis is not often available in EHPVO patients, it isbelieved that significant thrombosis of the main portal vein and its branchesoccurring at an early age leads to the development of this disease. Manyworkers'" are of the opinion that EHPVO and NCPF or IPH may berepresenting a spectrum of the same disease process, which has its origin inabdominal sepsis. The nature, extent, rapidity, host response and the age atthe time of portal vein infection probably influences whether the infectiveprocess would result into NCPF or EHPVO. Patients with distinct EHPVOmay have pyelephlebitic changes all along the portal vein and its branches.Even some hepatic histological features in EHPVO patients may resembleNCPF.4 On the other hand, patients with NCPF may also develop intra andextrahepatic portal vein thrombi. There is, however, as yet no conclusivehuman or experimental evidence to support the common origin hypothesis.

Diagnosis and differential diagnosis

For the physicians and the gastroenterologists practising in India, NCPF is adistinct entity and its diagnosis generally does not pose a problem. Threeconditions which, however, need to be differentiated from NCPF arecompensated cirrhosis in the young, EHPVO and tropical splenomegalysyndrome (TSS). Usually the history of jaundice, the clinical profile andraised transaminases indicate the presence of chronic liver disease.4" As theliver could be nodular and shrunken in NCPF, a liver biopsy to excludecirrhosis is mandatory. Occasionally laparoscopy may be of great help.Usually, EHPVO can be easily differentiated from NCPF as the patients aremuch younger and the portal vein block can be demonstrated with a SPV oran ultrasound examination. Tropical splenomegaly syndrome (TSS), whichresults from an abnormal and exaggerated immune response to chronicmalarial infection47 is another common entity seen in the tropics and Africa.It can easily be excluded because of the lack of any clinical signs of portalhypertension. The intrasplenic pressure may be raised in some, but wedgedhepatic pressure is within normal limits. "

411

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

It is rarely necessary to differentiate between splenomegaly of myeloidmetaplasia, Kala azar, Felty's syndrome with nodular regenerative hyper-plasia or idiopathic retroperitoneal fibrosis from NCPF.

Treatment and prognosis

At present, the 'pre-NCPF' or early stage of NCPF before the developmentof portal hypertension, is not known. The majority of the patients presentwith variceal bleeding and the treatment therefore involves control of acutebleeding or prevention of rebleeding by surgical or medicalmeans. Lienorenal shunt has been found to be quite useful. In a large series,mortality after shunt in patients with acute bleeds was reported to be 10%,but was negligible in elective shunt operations.4" The incidence of shuntocclusion, overt chronic portosystemic encephalopathy and rebleeding wasabout 10%. As well as the morbidity and mortality associated with surgery,the facilities and expertise for shunt surgery are not widely available. For thepast few years, many centres in India have started treating variceal bleeds inNCPF patients with endoscopic sclerotherapy. It has been found to be veryeffective in the management of acute variceal bleeding as well as preventionof rebleeding.49'" The incidence of variceal recurrence and rebleeding wereonly 22% and 3-1% respectively in a large series of patients.'4 At present,sclerotherapy is recommended as the treatment of choice for NCPF patientswith bleeding varices.

Although the natural history of NCPF is ill-understood, the disease seemsto have a protracted benign course with a very good prognosis."."l." Afterlienorenal surgery, a five year survival of 87% has been claimed.'" Aftersuccessful eradication of oesophageal varices with sclerotherapy, 100%survival has been reported in NCPF patients.'4 In fact, because the mostimportant cause of death in NCPF is an exsanguinating haemorrhage, thereseems a basis for evaluating prophylactic sclerotherapy in this group ofpatients.

In conclusion, NCPF should be understood as a common cause of portalhypertension in India characterised by repeated well tolerated attacks ofhaematemesis, long standing splenomegaly, near normal liver functions andno evidence of cirrhosis or portal vein obstruction. Although the aetiology isobscure, the pathological features of pyelephlebitis and obliterativeportovenopathy and some resemblance to EHPVO indirectly suggest a roleof portal vein infection. There is pathoanatomic and haemodynamicevidence of pre and perisinusoidal resistance to flow of portal blood withmarkedly raised intrasplenic and intravariceal pressures. The disease runs abenign course and once variceal bleeding can be successfully prevented, theprognosis is good. The fact that there are many similarities and somedifferences between NCPF of India and IPH of Japan indicates that thoughthe morphological response of the liver to different aetiological agents maybe similar, the clinical profile of the diseases could be quite different; youngmen in India and middle aged women in Japan.

S K SARINFrom the Department of Gastroenterology,G B Pant Hospital,New Delhi, india

412 Sarin

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

Non-cirrhotic portalfibrosis 413

Address for correspondence: Dr S K Satrin. Associate Professor. D)cpartrncnt of Giastroenterology. Gil Panlit liospital. NewDelhi. India.

Received for publication 28 July 1988.

References

1 Banti G. Splenomegalie mit Leberzirrhose. Beitrage zur pathologischen. Analt AllgemeinenPatliol 1889; 24: 21-33.

2 Ramalingaswami B, Wig KL. Sama SK. Cirrhosis of the liver in northern India.A clinicopathological study. Arch Intern Med 1962; 110: 350-8.

3 Subramanyam NT, Srinivasa Rao. Non-cirrhotic portal fibrosis. J Assoc Plhysicians India1970; 18: 935-40.

4 Basu AK. Boyer Mullick KC. Sen Gupta KP. Non-cirrhotic portal fibrosis with portalhypertension. Clinical and functional studies with results of operation. Indian J Med Res1967; 55: 336-51.

5 Proceedings of thze worksltop on non-cit-rrhlotic po)rtal fibro.sis. Indian Council of MedicalResearch, New Dehli 1969.

6 Boyer JL, Sengupta KP, Biswas SK, etal. Idiopathic portal hypertension. Comparison withthe portal hypertension of cirrhosis and extrahepatic portal vein obstruction. Ann InternMed 1967; 66: 41-68.

7 Tisdale WA, Klatskin G, Gleen WW. Portal hypertension and bleeding varices. Theiroccurrence in the absence of both intrahepatic and extrahepatic obstruction of the portalvein. N EnglJ Med 1959; 261: 209-18.

8 Mikkelsen WP, Edmondson HA, Peters RL, Redeker AG, Reynolds TB. Extra andintrahepatic portal hypertension without cirrhosis (hepatoportal sclerosis). Ann Surg 1965;162: 602-20.

9 Fogarty International Center. Diseases of the liver and biliary tract. Standardization ofnomenclature, diagnostic criteria and diagnostic methodology, Fogarty InternationalCenter, Washington DC 1976.

10 Okuda K. Idiopathic portal hypertension. In: Recent advances in heputology. Thomas HC,Jones EA, Edinburgh. Churchill Livingstone, 1986; 93-108.

11 Habibullah CM, Rao GN, Murthy DK, et al. Non-cirrhotic portal fibrosis in AndhraPradesh. J Assoc Phiysic ians Ind(ia 1978; 26: 379-82.

12 Koshy A. Rclationship between NCPF and EHO. In: Okuda K, Omata M, eds. Idiopathicportal hypertension. Tokyo: University of Tokyo Press, 1983: 13-7.

13 Sama SK, Bhargawa S, Gopi Nath N. etal. Non-cirrhotic portal fibrosis. Am J Med 1971; 51:16(0-9.

14 Sarin SK, Sachdev G, Nanda R. Follow-up of patients aftcr variceal cradication. Acomparison of paticnts with cirrhosis, non-cirrhotic portal fibrosis and extrahepaticobstruction. Ann Surg 1986; 204: 78-82.

15 Okuda K, Kono K, Ohnishi K, et al. Clinical study of eighty-six cases of idiopathic portalhypertension and comparison with cirrhosis with spienomcgaly. Gastr-oetnterology 1984; 86:600-10.

16 Eckhauser FE, Appleman HD, Knol JA, Strodal WE, Coran AG, Turcotte JG. Non-cirrhotic portal hypertension: differing patterns of diseasc in children and adults. Surgery1983; 94: 721-8.

17 Morris JS, Schmid M, Newman S. et al. Arscnic and non-cirrhotic portal hypertension.Gastroenterology 1974; 66: 86-94.

18 Datta DV, Mitra SK, Chuttani PN, Chakravarti RN. Chronic oral arsenic intoxication as apossible ctiological factor in idiopathic portal hypertension (non-cirrhotic portal fibrosis) inIndia. Gut 1979; 20: 378-84.

19 Narang APS, Dutta DV. Arsenic and liver. J Assoc Physicians Ind(ia 1983; 11: 87-90.20 Kono K, Ohnishi K, Omata M, et al. Experimental portal fibrosis produced by intraportal

injection of killed non-pathogenic Escherichia coli in rabbits. Gastroeteleology 1988; 94:787-96.

21 Mahajan RC, Datta DV. Malarial antibody studics in non-cirrhotic portal fibrosis -a preliminary observation. Indian J Med Res 1976; 64: 652-5.

22 Sharma BK, Malhotra P, Sarin SK, Kumar R, Nayyar AK, Broor SL. Malarial antibodystudies in patients with non-cirrhotic portal fibrosis. Proceedings of 28th AnnualConfcrcnce of Indian Society of (iastroenterology. Bhopal: 1987: B5-5.

23 Sarin SK, Babu S. Non-compressive myclopathy following lieno-rcnal shunt in a paticntwith non-cirrhotic portal fibrosis. Am J Gastroenterol 1983: 78: 440-1.

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

414 Sarin

24 Nayyar AK, Sharma BK, Sarin SK, Broor SL. Characterization of peripheral bloodlymphocytes in patients with non-cirrhotic portal fibrosis - a comparison with cirrhosis andhealthy controls. Proceedings of 28th/ Annual Conference of Indian Society of Gastno-enterology, Bhopal: 1987: B-3.

25 Malhotra P, Sharma BK, Nayyar AK, Sarin SK, Broor SL. Serum immunoglobulinscomplement and circulating immune compiexes in patients with non-cirrhotic portal fibrosis- a comparison with cirrhotics. Proceedings of 28th annual conference of Indian Society ofGastroenterology, Bhopal: 1987: B-2.

26 Warren KS. The immunopathogenesis of hepatosplenic schistosomiasis: from man tomonkey to molecule. In: Popper H, Schaffner F, eds. Progress in liver disease. VI. NewYork: Grune and Stratton, 1979:439-55.

27 Sarin SK, Malhotra V, Mehra NK, Anand BS, Taneja V. Familial aggregation in non-cirrhotic portal fibrosis - a study of four families. Am J Gastroenterol 1987: 82: 1130-4.

28 Taneja V, Mehra NK, Sarin SK, Vaidya MC. HLA studies in non-cirrhotic portal fibrosis.Tissue Antigen 1987; 30: 184-7.

29 Saha S. Portal venous and hepatic arterial blood flow in non-cirrhotic portal fibrosis, MDThesis, PGI, Chandigarh: 1974.

30 Nayak NC. Pathology of non-cirrhotic portal fibrosis in India. In: Okuda K, Omata M, eds.Idiopathic portal hypertension. Tokyo: Univ Tokyo Press, 1983: 37-47.

31 Aikat BK, Bhusnurmath SR, Chuttani PN, Mitra SK, Dutta DV. The pathology of non-cirrhotic portal fibrosis. Hum Patliol 1979; 10: 405-18.

32 Nayak NC, Ramalingaswami B. Obliterative venopathy of the liver. Alcli Patliol 1969: 87:359-69.

33 Okuda K, Nakashima T, Okudaira M, et al. Liver pathology of idiopathic portalhypertension. Comparison with non-cirrhotic portal fibrosis of India. Liver 1982: 2: 176-92.

34 Tandon BN, Lakshminarayan R, Bhargawa S, Nayak NC, Sama SK. Ultrastructure of theliver in non-cirrhotic portal fibrosis with portal hypertension. Gut 1970() 11: 905-10.

35 Sarin SK, Sethi KK, Nanda R. Measurement and correlation of wedged hepatic,intrahepatic, intraspienic and intravariceal pressures in patients with cirrhosis of liver andnon-cirrhotic portal fibrosis. Gut 1987; 28: 26(06.

36 Basu AK, Guhary BN. Non-cirrlhotic portal fibrosis wit/i portal hypertension. Calcutta:Academic Publications, 1978.

37 Sarin SK, Sethi KK. Tropical splenomegaly syndrome vs non-cirrhotic portal fibrosis - isthere a hemodynamic correlation. Hepatology 1988; 8: 1459.

38 Mohan JC, Sarin SK, Sethi KK, Nanda R, Anand BS, Arora R. The use of two-dimensionalcontrast echocardiography in patients with portal hypertension for detection of cardiacchamber abnormalities and pulmonary arterio-venous shunts. J Gastroenterol Hepatol1986; 1: 457-62.

39 Tandon BN, Nundy S, Nayak NC. Non-cirrhotic portal hypertension in Northern India:Clinical features and liver function tests. In: Okuda K, Omata M, eds. Idiopathic portalhypertension. Tokyo: Univ Tokyo Press, 1983: 377-86.

40 Sarin SK, Bihari M, Gupta M. Cerebral function in unoperated patients with non-cirrhoticportal fibrosis - a comparison with compensated non-alcoholic cirrhotics. J GastroenterolHepatol 1987; 2: 9-12.

41 Sarin SK, Nundy S. Subclinical encephalopathy after porta-systemic shunts in patients withnon-cirrhotic portal fibrosis. Liver 1985: 5: 1426.

42 Dutta DV, Agarwala ML, Narang APS, Samanta AKS. Chuttani PN. Ammonia tolerancetest in liver disease with special reference to non-cirrhotic portal fibrosis. Indian J Med Res1979: 70: 825-32.

43 Gaur SK, Vij JC, Sarin SK, Anand BS. Gastric secretory studies in cirrhosis and non-cirrhotic portal fibrosis. Digestion. 1988: 39: 151-5.

44 Sarin SK, Lal K, Sachdev G, Nanda R, Tandon RK. Biliary lipid composition in patientswith non-cirrhotic portal fibrosis - a comparison with compensated cirrhosis of liver. Liver1987; 7: 38-42.

45 Arora R, Mohanty M, Nundy S, Nayak NC. Phlebothrombosis as a common pathogenicdenominator in non-cirrhotic portal fibrosis and extrahepatic portal splcnic venousobstruction. Indian J Med Res 1984; 79: 392-403.

46 Sarin SK. Chari S, Sundaram KR, Ahuja RK, Anand BS, Broor SL. Young vs adultcirrhotics: A prospcctive, comparative analysis of the clinical profilc, natural coursc andsurvival. Gut 1988; 29: 101-7.

47 Hoffman SL. Piessens WF, Ratiwayanti S, et a1l. Rcduction of suppressor T lymphocytes inthe tropical splcnomcgaly syndromc. N Engl JMled 1984: 310: 337A41.

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from

Non-cirrhotic portalfibrosis 415

48 Nundy S, Tandon BN. The proximal lieno-renal shunt in the management of varices. In:Okuda K, Omata M, eds. Idiopathic portal hypertension. Tokyo: Univ Tokyo Press, 1983:535-44.

49 Sarin SK, Nanda R, Sachdev G, Chari S, Anand BS, Broor SL. Intravariceal versusparavariceal scierotherapy - a randomised, prospective, controlled clinical trial. Gut 1987;28: 657-62.

50 Sarin SK, Nanda R, Gaur SK, Nirmal K, Vij JC. Anand BS. Repeated endoscopicsclerotherapy for active variceal bleeding. Ann Surg 1985; 202: 708-11.

on Novem

ber 25, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.30.3.406 on 1 March 1989. D

ownloaded from